Gen AI Dashboard

  1. Home
  2. Gen AI
  3. Dashboard
AI WORKSPACE

One workspace for every AI workflow

Connect models, knowledge, and tools in one place. Give each team a clear path from prompt to reliable output.

All systems operational18 apps · 3 model routes · 362 sources
Open assistant Create with AI
AI assistant robot illustration
Applications18 active
Model routing3 endpoints
Knowledge362 sources

Cost & value

$12,37,842 of $50,00,000 budget
3.8%vs. last month
Verified savings$124from cached context
Budget remaining$1,15823% available

AI requests

Automated model calls
248.6k18.4%
Compared with previous 30 days
View request activity

Tokens processed

Input and output combined
18.4M12.1%
Compared with previous 30 days
Explore model usage

Median latency

P50 across active models
842 ms9.6%
Faster than previous 30 days
View response times

Quality score

Evaluation pass rate
94.8%2.3 pts
Above the 90% quality target
Review evaluations

AI request pipeline

Requests moving through each completion stage
Live
Stage conversion Counts for the last 30 days

Model usage

3 active
Requests
Claude Sonnet32%
79.6k requests · 940 ms
Metra GPT48%
119.3k requests · 612 ms
Gemini Pro20%
49.7k requests · 780 ms

Recent inference runs

Latest model executions across your AI agents
Run Agent Model Duration Tokens Status Cost
#RUN-84021Customer email summary Support copilot MetraUI GPT 724 ms 1,284 Completed $0.012
#RUN-84020Quarterly report draft Report writer Claude Sonnet 1.42 s 4,820 Completed $0.038
#RUN-84019Product description rewrite Commerce writer Gemini Pro 986 ms 962 Completed $0.009
#RUN-84018Policy question response Support copilot MetraUI GPT 2.08 s 2,106 Review required $0.021
#RUN-84017Lead call notes extraction Sales assistant Claude Sonnet 1.16 s 1,772 Completed $0.017
Showing the 5 most recent runsUpdated just now

Active model fleet

Endpoint health and traffic share
All healthy
MetraUI GPT Primary99.98% uptime · 48% traffic
612 msP50 latency
Claude Sonnet99.95% uptime · 32% traffic
940 msP50 latency
Gemini Pro99.91% uptime · 20% traffic
1.1 sP50 latency

Human review queue

Escalated outputs awaiting validation
8 pending
Waiting for review8
2 urgentOldest: 24 min
Safety filter blocked 14 outputs this week

Model response time

Compare median and tail latency across active models
Overall median842 ms9.6% this month
P95 latency1.86 sAcross all models
Under 2s SLA96.8%Target performance
Fastest model: MetraUI GPT · 612 ms medianLatency shown in milliseconds

Token efficiency

Input, output, and cached context
Total tokens18.4M
12.1%
Input tokens
7.2M 39%
Output tokens
7.3M 40%
Cached context
2.8M 15%
Reasoning tokens
1.1M 6%
$124 saved with caching
Estimated savings this month